The car sequencing problem (CarSP) was seen as a challenge to artificial intelligence. The CarSP is a version of the job-shop scheduling problem, which is known to be NP-complete. The task in the CarSP is to schedule a given number of cars (of different types) in a sequence to allow the teams in each workstation on the assembly line to fit the required options (e.g. radio, sunroof) on the cars within the capacity of that workstation. In unsolvable problems, one would like to minimize the penalties associated with the violation of the capacity constraints. Previous attempts to tackle the problem either have been unsuccessful or have been restricted to solvable CarSPs only. In this paper, we report on promising results in applying a generic genetic algorithm, which we call GAcSP, to tackle both solvable and unsolvable CarSPs.