The mouse limb bud micromass assay is one of many short‐term tests proposed as preliminary screens for potential developmental toxicity. Previous efforts to validate this assay have used too few “nonteratogens.” The purpose of this study was to examine additional compounds, most of which, based on the literature, were perceived to have low potential for developmental toxicity in vivo. In addition, a method of data analysis was sought that would identify selective developmental toxins in the micromass assay, i.e., those that are effective at dosages not maternally toxic. The concentration of each of 23 compounds that produced a 50% inhibition (IC50) of radiolabeled thymidine (T) and sulfate (S) incorporation was determined and used to calculate a T/S ratio. The T/S ratio may be a useful measure of developmental hazard, since T incorporation measures toxicity toward a general cell function (DNA synthesis) and S incorporation measures mainly toxicity toward a developmentally specific cell activity (chondroitin sulfate synthesis). All compounds tested produced T/S ratios of <2.0. Since 22 of these 23 compounds are classified as “nonteratogens” or nonselective developmental toxins in vivo, a low T/S ratio in this in vitro assay system may be capable of discriminating potential for developmental hazard in vivo. Copyright © 1990 Wiley‐Liss, Inc., A Wiley Company