This paper is concerned with the problem of measuring market response to a "communications mix"--the various means which a firm employs to transmit sales messages to potential buyers. Distributed lag models are applied to time series data for an ethical drug to estimate the short-run, intermediate, and long-run effects on market share of expenditures made for journal advertising, direct mail advertising, and samples and literature. Important differences were found among the communications variables with respect to the magnitude and over-time pattern of effect each had on market share. Significantly, the results indicate that the average historic allocation of resources to alternative communication vehicles has been in inverse relation to measured market response. A variety of econometric checks on the models lends credibility to the results. The managerial implications of the findings are discussed.