The choice of fabric should be taken seriously because it can make the difference between something that looks period and a costume fit only for Halloween. It's not even primarily about natural fibre vs. man-made fibre: Some man-made fibres, e.g. high-quality acetate (silk substitute), can fool even an experienced eye, whereas natural fibres can look totally off due to wrong colour and/or pattern.
I strongly recommend that you start by reading the page on Historically Correct Fabrics to see what's what for all periods prior to the mid 19th century. The rest of this page is specifically for the 18th century, although most of it is probably true for the Early Modern period.
Above: Two pairs of stays of Lampas liséré, last quarter 18th c
linen lining, last quarter 18th c
Indienne, late 18th
Linen is the traditional fibre for things that were routinely laudered, such as undergarments, kerchiefs (hand and neck), aprons, table and bed linens. White linen was associated with cleanliness, and more expensive because of the added effort of bleaching. Outer garments, such as aprons, pinafores or neckerchiefs, may also be dyed or woven in a check or stripe pattern.
From time to time the question is raised whether linen is also suitable for outer garments. It is difficult to answer because there are very, very few extant linen outer garments. It's hard to tell whether this is because they were rare to begin with, or because they were destroyed through second-, third- and fourth-hand use. The water-absorbent qualities of linen, which are enhanced by frequent laundering, make it uniquely suitable for cleaning and monthly purposes. The extant upper garments made from linen that I know of are a white quilted sack back gown and boy's suit in the Germanisches Nationalmuseum, a white quilted man's waistcoat (IIRC in the MetMuseum) and a very crudely made blue man's coat, probably working class, in Ludwigsburg. I also know of men's smocks, kind of like an oversized shirt, made of blue or white-and-blue striped linen in some regions of Germany.
Personally, I think that linen was used for outer garments of the (physically) working class because it was cheap, hard-wearing and a kind of wearable air condition in summer. I don't have proof, though. The white, quilted ones probably only "work" for upper class garments because they are (a) white and (b) quilted, i.e. until the advent of proof to the contrary, I shall regard non-white, non-quilted middle or upper class linen garments as non-existing.
Cotton was put to the same uses as linen, i.e. things that were laundered frequently. It feels less clammy in cold weather than linen does, which probably accounts for some of its early popularity. Apparently it is also a bit easier to dye than linen the earliest European attempts at dyeing multicolour floral patterns were done with cotton in the very late 17th century. You will often find it stated that cotton was forbidden e.g. in England, France and Prussia, but I have found that at least the English and French laws were much more specific: They prohibited the import of printed cotton fabrics, i.e. chintz, and the domestic production thereof. This would mean that white, striped and checked cotton fabrics were legal - and they were produced. In fact, even chintzes were produced in France despite the prohibtion1.
I've also spent some time researching the Proceedings of the Old Bailey and found that a) cotton garments were stolen in large numbers, which means that they were worn despite the ban, and b) that the values listed were not much higher than those of linen or woolen garments. This is only valid for England, but it shows that an official ban doesn't necessarily mean all that much.
Anyway, I know for a fact that cotton was merrily used in the Netherlands, Hamburg, Augsburg, Ulm, Leipzig and North America, where it was produced locally. The first three of those place even had their own chintz printing manufactories.
So the authentic fabrics for the 18th century are linen, wool, silk, cotton with caution, hemp and a fabric made from nettles which almost but not quite impossible to get nowadays.
Be careful about silks that have visible slubs. That would have been regarded as inferior quality and therefore used for lining, although a few outer garments made from slightly slubby silk exist. That is not, however, the same as modern-day dupioni! Not every slubby silk is dupioni and not every dupioni is slubby. Other types of silk one should stay away from are tussah, bourette, crêpe satin, crêpe de chine, pongée and stonewashed silk. Some people like to use chiffon and organza for kerchiefs and caps to make up for the lack of really fine linen, but that's not really authentic because caps and kerchiefs were meant to be washable at least, thinking of some 1780s cap creations, in theory.
For garments, the most suitable silk fabrics are taffeta, faille, duchesse satin, damask, brocade, gros de tours, lampas, moiré. They should be stiff and not too lightweight. And never, ever, pre-wash silk.
The three classic weaves - atlas (aka satin), plain (aka tabby aka linen) and twill - were all out there. Silk had the most fancy variations, such as lampas liséré, moiré and damask, velvet and corduroy. Of the latter I'm told that it was different from what you get now, but I've never seen the real thing up close. Linen was most often woven in linen weave, of course. I have found some linen twills as base and/or lining in garments, but not for any other use. The Frauenzimmer-Lexicon of 1715 mentions linen damask for bed and table linens. No cotton twill, or linen or cotton atlas so far. Wool did come in tabby and twill. Not sure about wool atlas, but apparently, wool damasks and brocades (!) existed2.
Not only the choice of fibre is governed by what a person could afford, but also the range of available colours. Some were easy to dye, some not so much.
The most important thing to remember about colours and that applies to all periods up until the mid-19th century is that 1., the only available dyes were plant (rarely animal) based and 2., that most such dyes work much better on animal fibres than on plant fibres. Linen in particular is difficult to dye. Only two types of dye work well with linen and cotton: indigo, because the dyeing process is completely different, and dyes with high tannin content, e.g. various tree barks, which dye varying shades of brown. And that is, basically, the extent of easily achievable colours in linen. I'm not saying they were the only ones technically possible 18th century dyers, like all 18th century artisans, were Cunning Artificers, after all , but they were the only likely ones if you go with the theory that linen wasn't "good" enough for the outer garments of anyone but the physically working class. So maybe the working class spent a little more for their sunday best but would Sunday best also be made of linen?
Cotton was easier to dye, but something like the fabled Turkish Red still required many more preparation steps 26 days all in all3 than wool or silk. Since cotton tended to be more expensive than linen, it was more likely to receive that extra effort, which I imagine was preferably spent on "nice" colours. Well, red. And, of course, red. (All the period dyeing books I've browsed seemed to be preoccupied with red more than anything else.) I have not yet seen any garment made from solid-coloured, plain-weave cotton, though. It seems that the dyers' efforts mostly went into the production of colourful chintzes (aka indiennes) in the Indian style, or into the production of dyed yarn for weaving striped and checked fabrics. My answer to the oft-heard question whether solid coloured cotton was authentic for outer garments is thus: Can your persona afford dyed cotton? Yes? Then why stop at cotton? Wear silk or at least indienne!
For wool and silk, practically anything is technically possible. Well, except day-glo. A brilliant and even green was pretty difficult to achieve and required at least two dye baths, i.e. two times the labour and cost of dye plants, makes for a more expensive fabric. A really deep, proper black even required up to three extremely saturated dye baths. I almost forgot to mention that the cost of dyeing was also higher the deeper, darker, more saturated a colour was. I said that blue on linen was easily achieved and therefore relatively cheap but navy or midnight blue required large amounts of indigo and lots of dips in the dye bath and therefore wasn't all that cheap after all.
I used to list a couple of (mostly pale) colours that seemed to be most popular, but after having seen a few thousand extant garments and portraits, I'd say that those who could afford it wore any colour they damn well pleased. OK, so I have yet to see a solid orange or violet gown or suit, and not even the best dyer could achieve "electric blue" and "electric purple", the first chemical dyes. The list of colours that apparently just weren't done is a lot shorter and harder to populate than the list of "seen that in museum XY".
There are a couple of "but"s I often hear when it comes to colours.
I guess most people who start making and/or wearing 18th century garments would prefer patterned fabrics. Because it's more colourful, but also because of a vague feeling that floral patterns are somehow typical of the period. But the more you learn about 18th century clothing, the more apparent it should become that flowers are by no means ubiquitous, and that there are lots and lots of traps. Telling suitable fabrics from unsuitable ones takes years of practice.
I have therefore created a page with good and bad examples.
When looking at patterns, it is extremely important to note not only the motifs and colours, but also the fibre. Patterns that are found on silk are not found on cotton, linen or wool. You will sometimes find patterns resembling those of (cotton) chintzes on silk, but something like the two examples of silk lampas above will not be found in cotton. The reasons are mostly of a technical (what was said about dyeing above) or economical nature (the relative value of each fibre, and thus, its show-off capabilites).
Silk takes dye easily, is for outer garments and is valuable. Patterns were woven, not printed, and very rarely (on Asian import silk), hand-painted4. In the period, it came as moiré, one- and multicolour damask, brocades on a basis of plain or patterned atlas, plain weave or faille, stripes, checks and ikat. Only two types of patterned silk are still available nowadays: same-colour damask and lampas liséré, the latter in the lower three-figure range per metre. Very rarely, moiré, and even more rarely, ikat (aka chiné). None of them is cheap.
Wool, while easily dyeable, usually wasn't patterned in the way that modern wool fabrics are, as e.g. herringbone, diamond twill or various checks. However there is evidence of woolen damasks and brocades2.
Is it cotton? Floral cotton prints (indienne, aka chintz) were very popular. Please remember the restrictions mentioned above: Blue and brown were easy, red and its derivatives (pink, orange) could be done, yellow wasn't popular, but dabbed into some indienne patterns. That's why the foliage on authentic indiennes is blueish rather than green: With the contemporary block printing technique, proper green would have meant to print yellow over the indigo blue, requiring a very good aim. If you look closely at period cotton prints, you'll see that the colours are mostly red and blue. Some violet, orange and pink could be achieved by using different mordants, but they're still created in the same red dye bath as the proper reds.
The Metmuseum has created a video that shows the process. It should explain better than words can why the colours and the styling of the flowers are the way they are. It should also tell you why period indienne patterns have no shading (i.e. colour bleeding from dark to light) and not a terrible lot of detail, i.e. it's rather two-dimensional.
If you come across cottons with a white or ecru background and a pattern in only one colour (usually red, blue, brown, black or green) that looks like a drawing or etching of a landscape, people, animals and the like, that's toile de jouy. Authentic for the latter half of the 18th century but, alas, not for clothing.
To cut a long story short: If you are unsure about a pattern, don't. Have another look at paintings with the pattern filter on (I mean the filter in your brain) and notice that 18th century clothing was not predominantly patterned, but predominantly solid. If you need variety, go wild with some trim! If you must have pattern, look at the "good and bad" page, a page of the Greater Bay Are Costumers Guild about printed cottons (many links are unfortunately broken) and the website Het geheugen van Nederland enter "sits" as search term. In the Provence and the Netherlands, indiennes have passed into traditional costume, which is why Dutch museums have such a lot of them.
1) The Bibliothèque nationale der France has swatchbooks
dating to the 1730s, i.e. within the period of the prohibition, of striped and
checked cotton and cotton-linen fabrics, and swatches
if chintz from Marseilles.
2) Baumgarten, Linda. What Clothes Reveal 2002, p. 114
3) Vollständiges Färbe- und Blaichbuch zu mehrem Unterricht, Nutzen und Gebrauch für Fabrikanten und Färber. Ulm: August Lebrecht Stettin, 1780. available at Google Books.
4) e.g.. Historical Fashion in Detail, p. 66