To study the effect of selective and non-selective post-emergence herbicides on weeds and yield of cotton, a field experiment was conducted during two consecutive seasons of Kharif 2014 and 2015 at CCS Haryana Agricultural University, Hisar.The experimental field was pre-dominantly infested with natural population of jungle rice (Echinochloa colona L.) and carpet weed (Trianthema portulacastrum L.) to the extent of 79 and 21% in 2014 and 71 and 29% during 2015, respectively. Application of pendimethalin at 1.0 kg/ha supplemented with other two hoeings at 20 and 50 DAS or one hoeing and post-emergence application of quizalofop-p-ethyl at 60 g/ha or one hoeing and post-emergence application of propaquizafop-p-ethyl at 62.5 g/ha at 60 DAS caused significant reduction in density and dry wt. of weeds as compared to weedy check up to harvest in both the years. Protected spray of glyphosate (0.5%) integrated with pendimethalin or paraquat (0.3%) with  parthiobac Na fb quizalofop-p-ethyl being at par with three mechanical weedings (at 20,40 and 60 DAS) helped to significantly reduce the population and dry weight of weeds at 90 DAS over weedy check. Weeds throughout the crop growing season reduced seed yield by 49.9 and 47.2% during 2014 and 2015, respectively. During 2014, all the treatments involving directed spray of either glyphosate or paraquat caused 8.3 - 10% toxicity to cotton crop where in 2015 the toxicity in these treatments was 5-8%. In 2014, maximum WCE (96.9%) was obtained with use of pendimethalin fb directed spray of glyphosate but during 2015, it was 83.3 with application of parthiobac-Na fb quizalofop-p-ethyl fb directed spray of glyphosate.Pendimethalin fb parthiobac-sodium caused maximum uptake of nitrogen during 2014 and 2015, which was 23.37 kg/ha and 24.68 kg/ha, respectively.