A digital modulation method using Chirp-Slope Keying (CSK) is developed for coherent underwater acoustic communications. Effective signal detection is a critical stage in the implementation of any communications system; we will see that CSK solves some significant challenges to reliable detection. This thesis is primarily based on analyzing the effectiveness of CSK through simulations using Matlab's Simulink for underwater communications. The procedure begins with modulating a chirp's slope by random binary data with a linear-down-slope chirp representing a 0, and a linear-up-slope chirp representing a 1. Each received symbol is demodulated by multiplying it with the exact linear-up-slope chirp and then integrating over a whole period (i.e., integrate and dump). This slope-detection technique reduces the need for the extensive recognition of the magnitude and/or the frequencies of the signal. Simulations demonstrate that CSK offers sturdy performance in the modeled ocean environment, even at very low signal-to-noise ratio (SNR). CSK is first tested using the fundamental communication channel, Additive White Gaussian Noise (AWGN) channel. Simulation results show excellent BER vs. SNR performance, implying CSK is a promising method. Further extensive analysis and simulations are performed to evaluate the quality of CSK in more realistic channels including Rayleigh amplitude fading channel and multipath.