SQL & PL/SQL

I am attempting to write a bit of SQL that can be used to do a count of Requisition Values, in different value bands, e.g.:

0 - 250
251 - 500
501 - 1000
etc...

I have got this far:

SELECT   sum_value
       , COUNT(req_id) req_ct
    FROM (SELECT   prha.requisition_header_id req_id
                 , CASE SUM(prla.unit_price * prla.quantity)
                      WHEN 250
                         THEN '250'
                      WHEN 500
                         THEN '500'
                      WHEN 500
                         THEN '1000'
                      ELSE 'OTHER'
                   END sum_value
              FROM po.po_requisition_headers_all prha
                 , po.po_requisition_lines_all prla
             WHERE prha.requisition_header_id = prla.requisition_header_id
               AND prha.creation_date >= SYSDATE - 3
          GROUP BY prha.requisition_header_id
                 , prla.requisition_header_id) rt
GROUP BY sum_value;

SUM_V     REQ_CT
----- ----------
250            3
500            6
OTHER        726

But - it is not that great, in that it counts REQs with an exact value of 250,400,500 etc.

I am trying to work out the syntax to allow the CASE statement to work with ranges, but I can't work it out.

I have naively attempted this, for example:

WHEN > 250 AND 500

But that errors out, because I can see that the '>' needs to know what it is checking

Any help much appreciated.

Thank you