A model is developed of the relation between adoption of an innovation by firms and firm size. Decision making is represented as a group assessment of the perceived ratio between expected returns after succesful introduction, and the risk that implementation will fail or will last too long to makc adoption worth while. In view of differences in thc assessment of uncertainty and corresponding demands on the perceived return-to-risk ratio, among influencers of decisions in the firm, decision making is taken as stochastic, with a probability of adoption related to (objective) net benefits, and probability of non-adoption related to risk. The model gives a new way of looking at the fact that small firms lag behind in the adoption of new technology. In contrast with previous models proposed by David and Davies the explanation has nothing to do with there being some critical size below which adoption is not profitable. The present model explains the adoption lag for smaller firms as the result of expected returns being proportional to size while risk of failed implementation is independent of size. The model is tested and estimated empirically on data for the adoption of computers in small scale retailing, in the Netherlands.