Although actively protected, the earliest cotton squares can sometimes be lost as a result of insect damage or an environmental stress. The loss of these earliest squares could lead to increased vegetative growth that could perpetuate square shed or otherwise reduce productive attributes of the crop. Pix (mepiquat chloride) could reduce excess vegetative growth brought on by early square loss or, alternatively, reduce the cotton plant's capacity to compensate for non productive early fruiting positions. Our objective was to evaluate the effect of early square loss in increasing levels with different rates and application timing of Pix on vegetative growth, yield, yield distribution on the plant, and crop earliness. The variety, Deltapine 50 was planted at two locations in early to mid-May, 1992. Three levels of square loss (SL) (0%, 20%, or 40% of all squares at 10-14 d after pin head square) and three rates and timings of Pix (0, 4-4 oz/Ac applications weekly initiated at match head square (4-4), and 2-8 oz/Ac applications biweekly initiated at early bloom (8-2)) were factorially arranged in a RCB. Pix reduced vegetative growth (plant height) and 40% SL increased plant height at both locations. There was no SL by Pix interaction for plant height, however. A significant (P<0.1) SL by Pix interaction did exist for lint yield. Averaged across locations, lint yields were 89 lb/Ac and 51 lb/Ac lower than 0% SL for 20% SL and 40% SL treatments, respectively, when no Pix was applied. With Pix at the 4-4 rate, however, yields after 20% SL were 90 lb/Ac and 141 lb/Ac more than 0% SL and 40% SL, respectively. At the 8-2 rate, Pix increased yields at 40% SL 124 lb/Ac above 0% SL and 20% SL. Alternatively, this rate resulted in an average yield that was 92 lb/Ac less than when no Pix was applied after 0% SL. Much of the compensation for early square loss which was improved by Pix application occurred on the adjacent positions to those that had fruit removed, i.e. position 2 of the first four sympodia. Productivity of positions which developed later in the season tended to increase with square loss and Pix application. These trends supported yield responses although means were not significantly different between any treatment. Pix applications tended to improve the percentage of total lint pickable by 2,662 DD60's, although differences were generally not significant. The results of this study suggested that Pix could play a role in improving yield potential after early square loss, but the interaction between Pix rate and timing and the amount of square loss would make it difficult to determine an accurate application procedure.