This data set includes (1) Experimental measurements, and (2) Simulation data. The experimental data set shows raw measurements and important calculated parameters. The simulation data are for the NIST vapor-compression system model, CYCLE_D-HX, and show the input data and output results. The simulation inputs are based on the experimental measurements.CYCLE_D-HX is a semi-theoretical model that simulates performance of a vapor-compression cycle with forced-convection heat exchangers for specified temperature profiles of the heat source and heat sink. In this study, we validated CYCLE_D-HX using experimental measurements from a small (< 4 kW capacity) heat pump test apparatus operated in cooling mode. We also applied the model to simulate performance of selected refrigerants in a system with optimized refrigerant circuitries in the evaporator and condenser. The tested refrigerants included the medium-pressure refrigerant R-134a and candidate replacements with a lower global-warming potential (GWP): R-513A, R-450A, R-134a/1234yf/1234ze(E) (49.2/33.8/17.0 mass %), R-515B, and R-1234yf. We also tested high-pressure refrigerant R-410A and candidate replacements with lower-GWP: R-32, R-452B, and R-454B. The model generally agreed with experimental results, with COP and Qvol overpredicted by (0 to 3) % for the basic cycle, and by (0 to 5) % for the cycle with the liquid-line/suction-line heat exchanger (LLSL-HX). Simulations with equal compressor efficiency and optimized tube circuitry showed the COP spread among medium-pressure refrigerants could be reduced to 3 % with proper design, compared to (12 to 33) % from the experiments. In optimized systems, the high-pressure refrigerants? COP was (1 to 6) % higher than the COP of the medium-pressure refrigerants. The LLSL-HX improved performance of refrigerants with high molar heat capacity (here, the medium-pressure refrigerants) by (1.0 to 1.5) %.