Wastewater treatment performance of three pilot-scale packed-bed anaerobic digesters with walnut shell medium was researched for treating dairy freestall barn flushwater. Reciprocation mixing was evaluated as a means to lessen channelization in the media bed and to improve biogas production and organic matter removal at ambient temperatures. Reciprocation has been used in biological nitrogen removal systems to introduce air into the system to repeatedly oxygenate nitrifying biofilm along with mixing (Behrends et al. 2003), but the anaerobic systems benefit from mixing. Two tanks were used in each system, where one was full and one was empty at any given time. Water was repeatedly pumped from one tank to the other and back again (reciprocation). A key research objective was to determine the minimum reciprocation frequency (between 0-10 per day) while still maintaining moderate methane production and treatment performance. Broken walnut shells with a specific surface area of 360 m2/m3 were used as the packed media. Digester influent, which was pretreated to remove large solids, had the following characteristics: total solids (TS) of 5.5 g/L, volatile solids (VS) of 2.8 g/L, 5-day carbonaceous biochemical oxygen demand (cBOD5) of 800 mg/L, and chemical oxygen demand (COD) of 4340 mg/L. Average digesting liquid temperatures ranged from 14.1 to 23.6 °C. At 6-day theoretical hydraulic residence times (V/Q where V is Lliquid, which is volume of liquid occupying the digester pores, and Q is total daily influent flow) and 1 reciprocation per day, methane production was 0.060 ± 0.10 LCH4/Lliquid-day and at 10 reciprocations methane production 0.058 ± 0.14 LCH4/Lliquid-day (mean ± standard deviation of measurements over time). COD percent removals were both 51% at 6-day V/Q. Since multiple reciprocations did not appear to make a difference in methane production and treatment performance, fewer reciprocations were used in subsequent experiments. Higher flow rates were also used in subsequent experiments to accelerate sludge clogging and channelization in the walnut-shell bed and thereby allow detection of any advantage provided by reciprocation compared to an upflow reactor. At 0 and 1 reciprocations per day and 0.35 and 0.50-day V/Qs, respectively, methane production was 0.24 ± 0.08 and 0.23 ± 0.08 LCH4/Lliquid-day and COD percent removal was 17 and 22%. Over the study period of 226 days, walnut shell porosities decreased due to sludge accumulation from 0.68 and 0.64 (start-up or clean-bed) to 0.31 and 0.24 in the 1 and 0 reciprocation per day reactors. Sludge accumulation and channelization did not appear to be affected by reciprocation mixing on the scale of this study.