We present phase-resolved, high-resolution (1.3 angstrom) spectroscopy of the brightest near-synchronous polar BY Cam taken on two different occasions in 1998 and 1999. The first tomographic study of such a system reveals line emission spread out over a large velocity range forming a crescent at negative v(y) velocities in the Doppler maps. In contrast to the majority of synchronous AM Her systems there is only weak indication for the presence of a focused accretion stream. These two facts suggest that the majority of the matter is accreted via an extended curtain. Location and extent of the structure in the Doppler maps can be reproduced with a simple curtain model raised over a wide (similar to 180 degrees) range in azimuth implying that the ballistic stream stretches to a point far behind the white dwarf. In order to reach such small magnetospheric radii mass accretion rates a factor of 10 to 20 in excess of that normally seen in polars would be required. In addition to the curtain emission, the Balmer lines show a narrow emission line component likely originating from the heated side of the secondary star. Its velocity amplitude of 190 km s(-1) together with an illumination model of the secondary star suggests a rather heavy white dwarf of M-1 >= 0.8 M-circle dot and an inclination larger than i >= 40 degrees. Timings of this feature in the present and historical data unequivocally determine the orbital period and have been used to establish a high-precision, long-term ephemeris.