Supernova 1987A remains the most well-studied supernova to date. Observations produced excellent broad-band photometric and spectroscopic coverage over a wide wavelength range at all epochs. We model the observed spectra from Day 1 to Day 81 using a hydrodynamical model. We show that good agreement can be obtained at times up to about 60 days, if we allow for extended nickel mixing. Later than about 60 days the observed Balmer lines become stronger than our models can reproduce. We show that this is likely due to a more complicated distribution of gamma-rays than we allow for in our spherically symmetric calculations. We present synthetic light curves in UBVRIJHK and a synthetic bolometric light curve. Using this broad baseline of detailed spectroscopic models we find a distance modulus mu = 18.5 +/- 0.2 using the SEAM method of determining distances to supernovae. We find that the explosion time agrees with that of the neutrino burst and is constrained at 68 percent confidence to within +/- 0.9 days. We argue that the weak Balmer lines of our detailed model calculations casts doubt on the accuracy of the purely photometric EPM method. We also suggest that Type IIP supernovae will be most useful as distance indicators at early times due to a variety of effects.