Signal degradation due to clipping is a threat to the performance of both the forward and the return path that has been given too little attention by our industry. In the forward path, the performance of fiber-optic transmission systems for broadband services is often limited by statistical clipping of the RF signal at the laser source. This means that systems meeting NCT A distortion and noise standards may nonetheless exhibit unacceptable video performance. Furthermore, the nature of these impairments is very different for directly-modulated, as opposed to externally-modulated optical sources. We report first on video tests of DFB lasers, indicating the nature and cause of clipping related impairments. Based on subjective picture quality tests, we have defined a straightforward quantitative test for determining acceptable clipping performance of a transmitter. The results of a similar test program for externally-modulated sources used in 1550 nm systems are described and compared with DFB's. Criteria for successful downstream data transmission jointly on lasers with analog signals are reported, as are preliminary investigations of 12VDC vs 24VDC RF amplifiers. Upstream digital transmission tests that investigate the effects of clipping both in return amplifiers and lasers are discussed, as well. Quantitative tests will be proposed as supplementary recommended practices for the industry.