Ground observations of magnetic perturbations can be used to study the spatial and temporal development of magnetospheric disturbances, provided that the magnetic effects of the quiet time ionospheric currents, known as Sq, can be removed from the data. AT mid-latitudes, variability of Sq currents may produce magnetic perturbations comparable to the magnetic effects of the partial ring current on comparable time scales. Thus it is important to know the error associated with the quiet field used as a reference to obtain the magnetic effects of the ring current. To determine this error, we have performed a statistical study using 2 years of quiet data (AE<100 &ggr;), acquired at 27 mid-latitude magnetic observatories. We have empirically determined the average quiet magnetic field at each observatory, and the uncertainty of that field as a function of season and local time. Since the Sq currents are confined primarily to the dayside, the uncertainty of the Sq diurnal variation changes as a function of local time. During disturbed periods the quiet field is subtracted from the observations from a worldwide chain of mid-latitude magnetic observatories to reveal the magnetic effects of substorms and the partial ring current. To obtain a smooth local time profile of the magnetic disturbance, we produce a least squares fit to the resulting data values, using a series of cubic spline functions. Profile uncertainty is estimated to range from approximately ¿23 &ggr; near local noon, to ¿10 &ggr; near local midnight. At dusk the uncertainty is near ¿13 &ggr;. These uncertainties represent uncertainty in the quiet field. The uncertainty can be greatly reduced by examining the changes which occur in some interval of time. This is done by subtracting the local time profile obtained at time t1 from the profile obtained at time t2. For time intervals less than 3 hours the quiet day errors are generally strongly correlated, and the errors are eliminated by the subtraction. The error in the difference profile is typically about ¿6 &ggr;. Thus for individual events the local time profile can be used to obtain parameters such as the magnitude, extent, and central meridian, which accurately characterize the magnetic disturbance attributed to a particular current system, provided the duration of the event is less than 3 hours. The more general mid-latitude indices, such as Dst and the dawn-dusk asymmetry, are less accurate, but they are still useful, provided the disturbance is sufficiently large.