Born in Boston, Massachusetts, Gilbert was educated at Harvard and at Cambridge University, England, where he obtained his PhD in physics in 1957. He returned to America to take up an appointment in theoretical physics at Harvard. He changed to molecular biology in 1960 under the influence of James Watson and in 1968 became professor of molecular biology at Harvard. He was elected chairman of the department of cellular and developmental biology in 1987.

In 1961 Jacques Monod and François Jacob proposed a theoretical answer to one of the most pressing problems of molecular biology, that of genetic control. If the common bacteria Escherichia coli is grown in the presence of milk sugar (lactose) it will produce an enzyme, beta-galactosidase, to split it into its component sugars. However, if grown in the absence of lactose, the enzyme will not be produced. There must therefore presumably be a mechanism whereby the gene controlling the production of the enzyme can be switched on and off. Monod and Jacob proposed a detailed account of such a mechanism, part of which involves the existence of a repressor molecule, which could bind itself to the gene and switch it off in the absence of lactose. The lac repressor, as it was called, would be inactivated, thus switching the gene on, by an inducer molecule produced by the lactose itself.

Plausible and powerful though the Monod–Jacob model appeared, it was still only a model until the basic confirmation provided by the isolation and identification of the lac repressor was achieved. Gilbert began such a search in 1965. This was a formidable task as the repressor was known to exist in small quantities only; nor was its chemical nature known. Gilbert himself likened the task to isolating the neutrino.

By 1966, in collaboration with Benno Muller-Hill, Gilbert had devised an ingenious experimental procedure, known as equilibrium dialysis. They used a specially active inducer, isopropyl thiogalactoside (IPTG), discovered by Melvin Calvin. Cells of E. coli were ground up and placed in a bag with a cellular membrane, allowing the passage of water and IPTG molecules but excluding such larger molecules as proteins. The bag was then placed in water containing radioactive IPTG.

As IPTG can pass through the bag an equal concentration of the inducer should be achieved. But if IPTG should bind itself to the lac repressor inside the bag then it will be too large to pass freely through the bag membrane. Consequently the concentration of the IPTG bound to the repressor should start to build up inside the bag and, being radioactive, should be readily detectable. Eventually they were able to report a concentration of IPTG 4% greater inside the bag than out. This was enough to encourage Gilbert and Muller-Hill to proceed to the next stage of fractionating, purifying, and isolating the repressor. This proved more difficult than they had expected but in late 1966 they were able to report the existence of a large protein molecule, the lac repressor. The following year their Harvard colleague M. Ptashne obtained a similar result with the lambda phage repressor.

Gilbert has also developed techniques for determining the sequence of bases in DNA, which though similar to Frederick Sanger's method differs in that it can be applied to single as well as double-stranded DNA. It was for this work that he shared the 1980 Nobel Prize for chemistry with Sanger and Paul Berg.